I can't find any tests and benchmarking for the API project. In particular, I am interested to see tests and benchmarks for the remote context propagation and serialization. Lack of tests creates impediments for suggesting improvements. For example, I am eager to propose something for #407 but it's not easy, I can't be sure if my changes would work.
If there's no particular reason for those tests to be missing, I can add some. If that's ok, shall I create a new project like OpenTelemetry.Api.Tests?
Related question - I am not quite sure why all the tests for the OpenTelemetry.Tests project are in the Impl directory, is there a guideline I should follow when placing the test files?
Hi @alexeyzimarev
We plan on adding additional testing and benchmark projects over the coming weeks. We see these as a prerequisite to a beta release. We actively encourage contributions, so if you feel you want to help, we'd be very happy for you to help.
@MikeGoldsmith it would be great if you can give me some initial guidance in regards to my other questions, so I can put things in the right place.
@alexeyzimarev here are the benchmarks for API. You can expand this as you wish.
Note that context propagation (if you mean context propagation over the wire) is done by .NET and hard to test without making real outgoing HTTP requests.
I meant tests for Extract and Inject. Both methods are reasonably complex and I believe it鈥檚 possible to test those methods in isolation.
Most helpful comment